Adam Smith attributed the ability of one man to do the work of many to the ‘invention of a great number of machines which facilitate and abridge labour’. Writing at the dawn of the Industrial Revolution, Smith witnessed one of the most profound reorganisations of human capital in history: the transition to mechanised production. The only comparable shift was humanity’s earlier move from nomadic hunter‑gathering to settled agriculture during the Neolithic Revolution.
Unsurprisingly, this transformation provoked deep anxiety. Contemporaneous sources such as the Leeds Woollen Workers’ Petition of 1786 warned that new machinery would deprive workers of the ‘opportunity of getting a livelihood’. Fears of mass unemployment, immiseration and social collapse accompanied technological progress then just as they do now.
Today, it is artificial intelligence that has revived these concerns. Rapid advances in AI systems have reignited worries about job insecurity to such an extent that universal basic income (UBI) has re‑entered mainstream political debate. Some policymakers now frame UBI as a necessary response to a looming wave of technological unemployment.
As reported this week, investment minister Jason Stockwood suggested that UBI could be used to ‘soft‑land’ industries most exposed to AI‑driven disruption, potentially funded through a windfall tax on technology companies. Public opinion appears receptive: YouGov polling indicates that 46% of Britons support some form of UBI.
Unlike existing welfare programmes, UBI is unconditional: it is neither means‑tested nor contingent on employment status. It represents a direct, universal transfer from the state to every individual. No country currently operates a comprehensive UBI system, largely because of its sheer fiscal cost, but some have trialled it, including the Welsh Government from 2022 to 2025. One study estimates that providing an annual UBI of just over £11,000 per person in Britain would require a flat income tax rate of around 45%.
Advocates increasingly argue that AI changes this calculus, making UBI not only necessary but financially feasible. At present, most economic value – and therefore tax revenue – is generated by human labour. If artificial general intelligence were to dramatically increase output without a corresponding increase in human work, large economic surpluses could, in theory, be generated with far fewer workers. The tax base could then shift away from labour towards non‑labour rents, including returns on capital and highly productive AI systems themselves.
AI could also raise public‑sector productivity, particularly in areas such as healthcare and education, reducing the state’s expenditure burden. In combination, faster productivity growth and a rebalanced tax base may create the fiscal space required for a universal income.
Even on these optimistic assumptions, however, UBI would require the wholesale replacement of the existing welfare system. In contrast to the complex eligibility structure of Universal Credit, UBI’s often hailed simplicity is precisely why it cannot sit alongside today’s sprawling benefits architecture. Running a universal cash transfer in parallel with an already expensive system would be fiscally untenable, especially as welfare spending is projected to rise sharply by 2030.
In any case, if reform is the objective, a negative income tax offers a more credible alternative. It delivers targeted income support with similar administrative efficiency, at far lower cost, and with weaker disincentives to work than a universal payment.
In recent weeks, the CEOs of Anthropic and JPMorgan have both cautioned that AI could displace large numbers of workers, while Morgan Stanley research has suggested that the UK is now losing more jobs than it is creating due to artificial intelligence. Yet this sense of alarm feels like déjà vu. The Leeds woollen workers voiced precisely the same fears over two centuries ago.
The Industrial Revolution did disrupt their livelihoods and did lead to periods of higher frictional unemployment. But those effects were temporary. Within 50 years’ time, creative destruction had replaced obsolete industries with more productive ones. The exponential improvements seen in living standards, child mortality and many other areas over the past centuries are a direct consequence of the progress that started with the Industrial Revolution.
There is little reason to believe the AI revolution will be fundamentally different. The UK, which boasts the world’s third‑largest AI sector, has the means to benefit significantly from this transformation. Rather than fearing technological progress or retreating into permanently redistributive solutions, policymakers should focus on enabling adaptation, mobility and growth. We are well placed to reap the gains of this fantastic and revolutionising technology – let’s not convince ourselves otherwise.
Britain and America have spent a decade arguing about Nimbyism as if it were a housing problem. It is much bigger than that: our inability to build, in a broad sense, has grown into a source of geopolitical weakness. The economic argument has been made already. Growth controls add anywhere from 17% to 38% to […]
‘A Waiter in Paris’ is the story of a young Englishman who graduated from a good university in London with a humanities degree, trained and educated for a world that no longer exists. ‘All while I applied for other, more “prestigious” jobs. Jobs that didn’t appear to exist anymore.’ That’s how Edward Chisholm puts it […]
One doesn’t like to be a disappointment to one’s parents. But in (at least) one crucial regard, I’ve let my dad down. He is a steam train enthusiast, and I’ve never shown a blind bit of interest. So it was more with filial duty than pleasure that I took the old man to the Didcot […]
